ABSTRACT
An approach to word disambiguation is presented that involves the use of a small number of general inference rules in conjunction with a hierarchically organized conceptual memory. This is in contrast to a commonly used method that relies primarily on disambiguation rules stored in the dictionary definitions of ambiguous words. The disambiguation of very vague or general words is problematical for the latter method because of its reliance on lexically based disambiguation rules. However, because of its more integrated organization, the disambiguation method presented here does not suffer from these problems.
